Thanks to factors such as Labour announcing AI’s “vast potential” for transformation across the UK, the already healthy appetite to understand AI use cases amongst senior leadership is ramping up.
Marketing teams face a critical question: How can we use AI while maintaining quality, authenticity, and environmental responsibility?
At least, we hope you’re asking that question before diving straight in!
Using AI to transform and not trash your marketing is a balancing act.
Let’s look at the checks and measures needed to make sure you don’t trash your marketing ‘equity’ with AI across the core areas of AI marketing transformation.
THE ELEPHANT IN THE ROOM
AI for Content Development
Currently, the most common way we see AI used in marketing strategies is in the development and creation of content. However, there are numerous ways using it to just pump out content can be damaging to your brand. Here are some ways you should be using it to assist your strategy.
Ideation & Brainstorming:
Content Creation:
If you’re careful with your briefing, AI can help you draft blog posts, social media content, and other business collateral. Note that I say draft. It needs to be fed with brand guidelines, key messages, audience information, key research and internal expertise.
If you choose to use AI you need to be very well aware of the tell-tale signs of AI-generated content.
Once you know what to look for, it becomes surprisingly easy to spot. Here are some key markers that separate AI generated content from original work.
AI Generated Red Flags
- AI tends to use similar sentence patterns, including excessive use of "however," "moreover," "additionally", as well as repeated use of other common words. Leverage, anyone?
- Lots of bullet points & lists (oh, the irony)
- Lack of specific, nuanced industry knowledge or practical experience
- Consistent tone throughout: Humans naturally vary their tone; AI maintains mechanical consistency
- Oddly perfect grammar can be a red flag - humans occasionally break rules for effect
- Americanisms and/or US English if it’s written for a British English audience
- Lacking actual opinion!
- Unnaturally perfect keyword placement
- Repetitive themes
- Missing empathy
- Bland beige word salad resulting in a blog without a strong narrative
There is a better way
How to Use AI for Marketing in a Strategic Way
So, how can you use AI to get the productivity bump without falling foul to the pitfalls?
You need to reframe AI as a support tool, not a silver bullet for content creation.
Instead of using AI to generate content, use it to:
Click below to expand
e.g. Feed your AI with an industry report and ask it to pull out key takeaways. Or, give it some stats you’ve collected and ask it to build a blog outline around them, or even a case study. But remember this is your starting point.
The Human Trinity:
How To Use AI in Marketing Effectively
Your AI outputs will only ever be as good as the humans driving them. While the tools highlighted above have benefits, the healthiest way to view them is as amplifiers of human skill, not replacements for it.
The Strategy
Without a clear strategy, AI is just sophisticated guesswork. Your strategy must include:
- Clear business objectives that align with overall company goals
- Deep understanding of target audience needs and pain points
- Well-defined brand voice and positioning that distinguishes you from competitors
- Long-term vision alignment to ensure consistent direction
- Measurable outcomes that matter to your business
Resource Alert!
The Brief & Feed
Click below to expand
You can use transcripts of meetings to provide background information for your prompts.
Ensure you make clear any conventions of your industry or brand that people would expect to see in your content.
Clarify your company’s tone of voice. AI content does still struggle with this, but it will give you a good start.
If you can provide central themes and core messaging, there is less to build out later.
Everything is stronger with data. Make sure you provide any key statistics that will help drive the narrative (and it might stop any made-up statistics creeping in there!).
"Using AI to create a framework or structure is far more time-effective than writing full drafts."
The Edit
We can all tell when AI content has been plain copied and pasted – and so can Google. A great edit can transform AI-generated content from bland to brilliant.
BUT, you do still run the risk of missing weaknesses. This is why we favour using AI to support the framework & structure of a piece – rather than writing a full draft to then amend/edit, as it can feel like plastering over the cracks.
Here’s an example. We gave ChatGPT a brief for this blog, which included background information about TFG, our audiences, our tone of voice, and the intent of the blog. You can find the result of that here.
You can spot many of the AI red flags we mention above. In this case, there aren’t too many of the language-related ones, but the whole blog reads as bland and uninteresting. It repeats itself in most sections and there are no tangible examples. The blog generated would need a very heavy edit to get it to a similar level as this blog you’re reading now (not to toot our own horns).
How much time is that truly saving?
How to Edit AI Generated Content
- Audience-aligned. Does the brand write in US English or British English normally? Ensure it matches.
- Do away with very typical AI language.
- Un-corporate the language (if that’s a word).
- Add human opinion, and authoritative industry insight.
- Don’t do a Zuckerberg - fact check everything! Time and time again I come across AI generated content that contains simply made up references and statistics
- Run your finished piece through an AI checker such as Quillbot or ZeroGPT. These tools aren’t infallible but they are a good indicator of how much your content smacks of AI. As an example, our test piece above, which hasn’t been edited at all, scored 85-91%.
The Cost of Careless AI-Use
- Loss of authentic brand voice; you run the risk of sounding like everyone else (been on LinkedIn lately? Yup.)
- Disconnection from the target audience and so reduced customer trust
- Potential PR backlashes from AI-generated mistakes. Fact check, fact check, fact check
- Wasted time, energy and budget on ineffective content
AI vs the planet
Environmental and ethical Considerations
As a B Corp, this is a huge one for us.
In 2024 Google had to declare that their emissions had risen 48% in the last five years and 13% in the last 12 months, largely driven by increased data centre demand to service its AI needs.
And it’s not just emissions – building AI models demands significant water usage for data centre cooling to prevent overheating.
Each time someone uses ChatGPT, it generates between 2.5 to 5 grams of CO2, averaging about 4.32 grams per query. While a single query’s carbon footprint might seem minimal, the environmental impact becomes more notable when considering ChatGPT’s massive daily usage.
Impact of ChatGPT Queries
To make this more relatable, here’s how ChatGPT queries compare to everyday activities:
- Running ChatGPT 15 times produces similar CO2 emissions to watching an hour of video content
- Using ChatGPT 16 times is comparable to the emissions from boiling one kettle of water
- The CO2 from 20-50 ChatGPT queries equals the carbon footprint of producing half a litre of drinking water
Let’s put this into perspective with something we’re all familiar with.
Google processes around 3.5 billion searches every day, with each search generating approximately 0.2 grams of CO2. This adds up to 700 million grams of carbon dioxide daily – enough emissions to fuel seven round trips to the moon by car.
So, while using AI isn’t ideal – neither is your individual contribution to that global Google search pile-on.
Like everything, it’s all in moderation.
*Note: Environmental impact figures are based on current research and may vary depending on specific circumstances, technology improvements, and changes in energy grid composition.*
What's Next?
Looking to the Future
Keep an eye on Industry developments
Google hit the headlines with their wildly increased emissions. They’ve since made a deal to drive $20bn of renewable energy investment for powering its AI data centres, partnering with Intersect Power and TPG’s climate investment arm.
They’re effectively aiming to supply clean energy and storage solutions for new facilities, with Google adopting a “power first” approach to “completely rethink” how it develops data centres with the first project due for completion in 2027. Considering that global AI emissions are now reported to match that of the airline industry (Dr Van Rijmenam), you could say they’re shutting the door after the horse has bolted.
Balancing Impact and Innovation
Can you balance productivity, quality and environmental responsibility as you transform your marketing with AI?
Answer: not really. Yet. But if you are going to use it – at least be mindful. Remember, technology should serve your strategy, not define it.
If you’re after support with your marketing strategy (with or without AI!), do contact us.
